Share your thoughts, 1 month free Claude Pro on usSee more

Grounding on ScreenSpot v2 (test)

95.3Overall Accuracy

UI-Venus-Ground-72B

Updated 3d ago

Evaluation Results

Method	Links
UI-Venus-Ground-72B 2026.02		95.3	99.7	93.8	95.9	90	96.2	92.6
GUI-Owl-32B 2026.02		93.2	98.6	90	97.9	87.8	94.4	86.7
JEDI-7B 2026.02		91.7	96.9	87.2	95.9	87.9	94.4	84.2
K²-Agent 2026.02		91.3	96.9	80.6	95.9	83.6	95.3	90.6
UI-TARS-72B 2026.02		90.3	94.8	86.3	91.2	87.9	91.5	87.7
Claude 3.7 Sonnet 2026.02		87.6	-	-	-	-	-	-
LAMO-3B 2026.04		86.4	98.6	83.7	96.1	76.8	89.1	74.2
LAMO-3B 2026.04		86.4	98.6	83.7	96.1	76.8	89.1	74.2
HAR-GUI 2026.04		86.2	96.5	81	95.4	76.5	88.8	78.8
HAR-GUI-3B 2026.04		86.2	96.5	81	95.4	76.5	88.8	78.8
UI-R1 2026.04		85.4	96.2	84.3	92.3	63.6	89.2	75.4
UI-R1-3B 2026.04		85.4	96.2	84.3	92.3	63.6	89.2	75.4
UI-TARS 2026.04		84.7	95.2	79.1	90.7	68.6	90.6	77.3
UI-TARS-2B 2026.04		84.7	95.2	79.1	90.7	68.6	90.6	77.3
OS-Atlas 2026.04		84.1	95.2	75.8	90.7	63.6	90.6	77.3
OS-Atlas-7B 2026.04		84.1	95.2	75.8	90.7	63.6	90.6	77.3
InternVL3 2026.04		81.4	-	-	-	-	-	-
InternVL3-8B 2026.04		81.4	-	-	-	-	-	-
Qwen2.5-VL 2026.04		81.3	95	80.1	90.2	64.3	88	70.4
Qwen2.5-VL-3B 2026.04		81.3	95	80.1	90.2	64.3	88	70.4
GPT-4o + OS-Atlas-4B 2026.04		79.1	95.5	75.8	79.4	49.3	90.2	66.5
GPT-4o + OS-Atlas-4B 2026.04		79.1	95.5	75.8	79.4	49.3	90.2	66.5
OS-Atlas 2026.04		71.9	87.2	59.7	72.7	46.4	85.9	63.1
OS-Atlas-4B 2026.04		71.9	87.2	59.7	72.7	46.4	85.9	63.1
Operator 2026.02		70.5	47.3	41.5	90.2	80.3	92.8	84.3
GPT-4o + SeeClick 2026.04		63.6	85.2	58.8	79.9	37.1	72.7	30.1
GPT-4o + SeeClick 2026.04		63.6	85.2	58.8	79.9	37.1	72.7	30.1
SeeClick 2026.04		55.1	78.4	50.7	70.1	29.3	55.2	32.5
SeeClick 2026.04		55.1	78.4	50.7	70.1	29.3	55.2	32.5